← Back to arXiv
arXivAnalysis / PDEsarXiv:2607.18547

The Stefan problem for complete melting of finitely strained solids into viscoelastic fluids

A material like ice can melt into water and refreeze repeatedly. Capturing this process mathematically is surprisingly hard because the material changes its fundamental mechanical character: a solid resists being twisted or sheared elastically, while a fluid does not. This paper develops a unified mathematical framework that handles both states and the transitions between them within a single set of equations, all written from the perspective of a fixed observer in space rather than tracking individual material particles. The key physical idea is that the material's resistance to shearing depends on temperature and can fade away entirely as it melts, allowing the liquid to flow freely, refreeze into whatever shape it ends up in, and potentially melt again, cycling through these states indefinitely.

The mechanical description uses a model called Jeffreys rheology, which combines elastic springlike behavior with viscous fluid-like dissipation in a way that smoothly accommodates the solid-to-liquid transition. Heat exchange at the moving interface between solid and liquid is handled through a classical framework called the Stefan problem, which accounts for the latent heat absorbed or released during melting and freezing. The authors also include more realistic physical effects like kinetic overheating and undercooling, meaning the transition does not have to happen exactly at the melting temperature but can lag slightly depending on how fast the process unfolds.

To prove that their equations actually have solutions, the authors use a mathematical technique called time discretization, where the continuous time evolution is approximated by a sequence of small steps. They also introduce a regularization called a higher-gradient or multipolar continuum modification, which adds a little extra mathematical smoothness to keep the equations well-behaved. Together these tools allow them to rigorously establish that the problem is mathematically consistent, laying groundwork for future numerical simulations of real melting and freezing processes in deformable solids at large deformations.

Read original →